Author ORCID Identifier

Document Type


Date of Award


Degree Name

Doctor of Philosophy (PhD)


Basic Biomedical Science

First Advisor

Erik A Ehli


INTRODUCTION: Breast cancer (BC) is the most common cancer among women and is classified as a complex disease. Advances in population genomics have led to the development of polygenic risk scores (PRSs) with the potential to enhance current risk models, but replication is often limited. OBJECTIVE: We sought to assess the predictive capabilities of two high-powered BC PRSs in a sample population selected for breast cancer. In addition, the capacity of the PRSs to predict clinical variables that could improve BC screening and treatments was explored. METHODS: Two published PRS algorithms (313 vs 3820) were used to score female subjects in this retrospective case-control study utilizing genetically similar populations from the integrated Cancer Repository for Cancer Research and Netherland Twin Register as BC positive and negative subjects, respectively. All subject biospecimens were genotyped on a custom Illumina Global Screening Array followed by standardized quality control, imputation, and principal component analysis (PCA). Phenotypic data was collected using patient-based questionnaires for cases and controls, with additional electronic medical record data. 261 cases and 1,303 controls were scored and PRS performance and associations were compared using means testing and receiver operating characteristic curve (ROC) analysis. RESULTS: Mean PRSs were significantly different (p<0.001) between cases and controls for both PRSs (313_PRS: 0.4122 vs. -0.0236, 3820_PRS: 0.4393 and -0.0305, respectively). ROC analysis showed an area of 0.609 and 0.619 for the 313 PRS and 3820 PRS, respectively. Survival curve analysis revealed a increase in BC incidence in the highest 313_PRS bin starting at approximately age 40 that persisted throughout life. Odds ratio (OR) calculations revealed that the 313_PRS improved risk stratification across the lifespan and increased with age. Clinical phenotypes did not show significant associations. CONCLUSIONS: Our study provides further evidence of the reproducibility and predictive performance of two previously published BC PRSs utilizing an independent study population. Our analyses indicate that the 313_PRS is better able to stratify BC risk, especially for those over age 60. Associations of BC clinical phenotypes and PRS were not significant, indicating the specificity and limitations of PRS use and the need for trait specific PRS development.

Subject Categories

Biostatistics | Genetics | Medicine and Health Sciences


bioinformatics, breast cancer, genetic risk, polygenic risk scores, population genetics

Number of Pages



University of South Dakota

Available for download on Wednesday, August 28, 2024